[cfp-interest 3797] Re: Updates to Annex F Special Cases + 10.1

[Damian McGuckin]

On Fri, 27 Feb 2026, Paul Zimmermann wrote:

> I can access the documents after I login to WebHome.

Cool.

> C2Y-F-10-1-Only-20260227.pdf:
> F.10.1#3: maybe we can group f(x) and f(x,y) to avoid repetitions

I tried and was less than clear.

> F.10.1#14: floating point -> floating-point

Thanks.

> F.10.1#14: if the input is sNaN, it should raise invalid, no?

You need to read the changes to 2.2.

When the C standard says NaN, it implies qNaN. This has been the default 
situation long before out changes.  Please see

   https://cfp-wiki.esi.com.au/pub/CFP/WebHome/C2Y-F-2-2-Only-20260210.pdf

So F.10.1#14 never has sNaN input.

There is a discussion in C2Y-F-2-2... about what happens with sNaN input.

Thanks - Damian